ox 2.7.0 → 2.8.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 75309ac947bc1f273ca4f6f65096f8fd114bf4ab
-  data.tar.gz: 481e35498618a38fde11809e04abf896024ab43f
+  metadata.gz: af2271c3c5cd6947ce101436041312351fd0d4f5
+  data.tar.gz: c1589595a362f3fe1111fc633c33626e529c6792
 SHA512:
-  metadata.gz: 3fa20f3ffd22f25b93984ac9f89948a76f90b6398e085980b7295ed634e0e1baf4399c43e2612821da8f47dd2ea8c0da01ae9712923f7042ab646ac7c463c406
-  data.tar.gz: d2769754eea05431e2e0b6be87bc0bed35992d28ab0b159d967a2198dcaa44489d538f48d96de1a4d8152f2a72ad2d5563b580a2f9064281ef876eb55d9fb5d3
+  metadata.gz: 62b558973eb617a8cc52c44b222babfaf8315433385265deea21abca52cbd5994146befbfbb785bc3878663aa2e2fc053fe101bf7e0982c0464f06d9114e9b41
+  data.tar.gz: e86031b3635268f92f19f9e17f2a1619a0729e01c3bde9652f26097c24a3a9665335b4a49d6a5840ade33d7690d41c17a3d27088d08166b54e0aad1ca111d680

data/CHANGELOG.md

@@ -1,4 +1,9 @@
 
+## 2.8.0 - September 22, 2017
+
+  - Added :skip_off mode to make sax callback on every none empty string even
+    if there are not other non-whitespace characters present.
+
 ## 2.7.0 - August 18, 2017
 
   - Two new load modes added, :hash and :hash_no_attrs. Both load an XML

data/ext/ox/encode.h

@@ -0,0 +1,26 @@
+/* encode.h
+ * Copyright (c) 2011, Peter Ohler
+ * All rights reserved.
+ */
+
+#ifndef __OX_ENCODE_H__
+#define __OX_ENCODE_H__
+
+#include "ruby.h"
+#if HAS_ENCODING_SUPPORT
+#include "ruby/encoding.h"
+#endif
+
+static inline VALUE
+ox_encode(VALUE rstr) {
+#if HAS_ENCODING_SUPPORT
+    rb_enc_associate(rstr, ox_utf8_encoding);
+#else
+    if (Qnil != ox_utf8_encoding) {
+	rstr = rb_funcall(ox_utf8_encoding, ox_iconv_id, 1, rstr);
+    }
+#endif
+    return rstr;
+}
+
+#endif /* __OX_ENCODE_H__ */

data/ext/ox/ox.c

@@ -129,6 +129,7 @@ static VALUE	opt_format_sym;
 static VALUE	optimized_sym;
 static VALUE	overlay_sym;
 static VALUE	skip_none_sym;
+static VALUE	skip_off_sym;
 static VALUE	skip_return_sym;
 static VALUE	skip_sym;
 static VALUE	skip_white_sym;
@@ -278,7 +279,7 @@ hints_to_overlay(Hints hints) {
  * - _:mode_ [:object|:generic|:limited|:hash|:hash_no_attrs|nil] load method to use for XML
  * - _:effort_ [:strict|:tolerant|:auto_define] set the tolerance level for loading
  * - _:symbolize_keys_ [true|false|nil] symbolize element attribute keys or leave as Strings
- * - _:skip_ [:skip_none|:skip_return|:skip_white] determines how to handle white space in text
+ * - _:skip_ [:skip_none|:skip_return|:skip_white|:skip_off] determines how to handle white space in text
  * - _:smart_ [true|false|nil] flag indicating the SAX parser uses hints if available (use with html)
  * - _:convert_special_ [true|false|nil] flag indicating special characters like < are converted with the SAX parser
  * - _:invalid_replace_ [nil|String] replacement string for invalid XML characters on dump. nil indicates include anyway as hex. A string, limited to 10 characters will replace the invalid character with the replace.
@@ -330,6 +331,7 @@ get_def_opts(VALUE self) {
     default:			rb_hash_aset(opts, effort_sym, Qnil);			break;
     }
     switch (ox_default_options.skip) {
+    case OffSkip:		rb_hash_aset(opts, skip_sym, skip_off_sym);		break;
     case NoSkip:		rb_hash_aset(opts, skip_sym, skip_none_sym);		break;
     case CrSkip:		rb_hash_aset(opts, skip_sym, skip_return_sym);		break;
     case SpcSkip:		rb_hash_aset(opts, skip_sym, skip_white_sym);		break;
@@ -413,7 +415,7 @@ sax_html_overlay(VALUE self) {
  *   - _:mode_ [:object|:generic|:limited|:hash|:hash_no_attrs|nil] load method to use for XML
  *   - _:effort_ [:strict|:tolerant|:auto_define] set the tolerance level for loading
  *   - _:symbolize_keys_ [true|false|nil] symbolize element attribute keys or leave as Strings
- *   - _:skip_ [:skip_none|:skip_return|:skip_white] determines how to handle white space in text
+ *   - _:skip_ [:skip_none|:skip_return|:skip_white|:skip_off] determines how to handle white space in text
  *   - _:smart_ [true|false|nil] flag indicating the SAX parser uses hints if available (use with html)
  *   - _:invalid_replace_ [nil|String] replacement string for invalid XML characters on dump. nil indicates include anyway as hex. A string, limited to 10 characters will replace the invalid character with the replace.
  *   - _:strip_namespace_ [nil|String|true|false] "" or false result in no namespace stripping. A string of "*" or true will strip all namespaces. Any other non-empty string indicates that matching namespaces will be stripped.
@@ -503,6 +505,8 @@ set_def_opts(VALUE self, VALUE opts) {
     v = rb_hash_aref(opts, skip_sym);
     if (Qnil == v) {
 	ox_default_options.skip = NoSkip;
+    } else if (skip_off_sym == v) {
+	ox_default_options.skip = OffSkip;
     } else if (skip_none_sym == v) {
 	ox_default_options.skip = NoSkip;
     } else if (skip_return_sym == v) {
@@ -510,7 +514,7 @@ set_def_opts(VALUE self, VALUE opts) {
     } else if (skip_white_sym == v) {
 	ox_default_options.skip = SpcSkip;
     } else {
-	rb_raise(ox_parse_error_class, ":skip must be :skip_none, :skip_return, :skip_white, or nil.\n");
+	rb_raise(ox_parse_error_class, ":skip must be :skip_none, :skip_return, :skip_white, :skip_off, or nil.\n");
     }
 
     v = rb_hash_lookup(opts, convert_special_sym);
@@ -732,12 +736,14 @@ load(char *xml, int argc, VALUE *argv, VALUE self, VALUE encoding, Err err) {
 	if (Qnil != (v = rb_hash_lookup(h, skip_sym))) {
 	    if (skip_none_sym == v) {
 		options.skip = NoSkip;
+	    } else if (skip_off_sym == v) {
+		options.skip = OffSkip;
 	    } else if (skip_return_sym == v) {
 		options.skip = CrSkip;
 	    } else if (skip_white_sym == v) {
 		options.skip = SpcSkip;
 	    } else {
-		rb_raise(ox_parse_error_class, ":skip must be :skip_none, :skip_return, or :skip_white.\n");
+		rb_raise(ox_parse_error_class, ":skip must be :skip_none, :skip_return, :skip_white, or :skip_off.\n");
 	    }
 	}
 
@@ -996,7 +1002,7 @@ load_file(int argc, VALUE *argv, VALUE self) {
  *   - *:convert_special* [true|false] flag indicating special characters like < are converted
  *   - *:symbolize* [true|false] flag indicating the parser symbolize element and attribute names
  *   - *:smart* [true|false] flag indicating the parser uses hints if available (use with html)
- *   - *:skip* [:skip_none|:skip_return|:skip_white] flag indicating the parser skips \\r or collpase white space into a single space. Default (skip space)
+ *   - *:skip* [:skip_none|:skip_return|:skip_white|:skip_off] flag indicating the parser skips \\r or collpase white space into a single space. Default (skip space)
  *   - *:strip_namespace* [nil|String|true|false] "" or false result in no namespace stripping. A string of "*" or true will strip all namespaces. Any other non-empty string indicates that matching namespaces will be stripped.
  */
 static VALUE
@@ -1033,6 +1039,8 @@ sax_parse(int argc, VALUE *argv, VALUE self) {
 		options.skip = SpcSkip;
 	    } else if (skip_none_sym == v) {
 		options.skip = NoSkip;
+	    } else if (skip_off_sym == v) {
+		options.skip = OffSkip;
 	    }
 	}
 	if (Qnil != (v = rb_hash_lookup(h, strip_namespace_sym))) {
@@ -1069,7 +1077,7 @@ sax_parse(int argc, VALUE *argv, VALUE self) {
  * - +options+ [Hash] options parse options
  *   - *:convert_special* [true|false] flag indicating special characters like < are converted
  *   - *:symbolize* [true|false] flag indicating the parser symbolize element and attribute names
- *   - *:skip* [:skip_none|:skip_return|:skip_white] flag indicating the parser skips \\r or collapse white space into a single space. Default (skip space)
+ *   - *:skip* [:skip_none|:skip_return|:skip_white|:skip_off] flag indicating the parser skips \\r or collapse white space into a single space. Default (skip space)
  *   - *:overlay* [Hash] a Hash of keys that match html element names and values that are one of
  *     - _:active_ - make the normal callback for the element
  *     - _:nest_ok_ - active but ignore nest check
@@ -1113,6 +1121,8 @@ sax_html(int argc, VALUE *argv, VALUE self) {
 		options.skip = SpcSkip;
 	    } else if (skip_none_sym == v) {
 		options.skip = NoSkip;
+	    } else if (skip_off_sym == v) {
+		options.skip = OffSkip;
 	    }
 	}
 	if (Qnil != (v = rb_hash_lookup(h, overlay_sym))) {
@@ -1455,6 +1465,7 @@ void Init_ox() {
     ox_standalone_sym = ID2SYM(rb_intern("standalone"));	rb_gc_register_address(&ox_standalone_sym);
     ox_version_sym = ID2SYM(rb_intern("version"));		rb_gc_register_address(&ox_version_sym);
     skip_none_sym = ID2SYM(rb_intern("skip_none"));		rb_gc_register_address(&skip_none_sym);
+    skip_off_sym = ID2SYM(rb_intern("skip_off"));		rb_gc_register_address(&skip_off_sym);
     skip_return_sym = ID2SYM(rb_intern("skip_return"));		rb_gc_register_address(&skip_return_sym);
     skip_sym = ID2SYM(rb_intern("skip"));			rb_gc_register_address(&skip_sym);
     skip_white_sym = ID2SYM(rb_intern("skip_white"));		rb_gc_register_address(&skip_white_sym);

data/ext/ox/ox.h

@@ -96,6 +96,7 @@ typedef enum {
 } LoadMode;
 
 typedef enum {
+    OffSkip  = 'o',
     NoSkip   = 'n',
     CrSkip   = 'r',
     SpcSkip  = 's',

data/ext/ox/parse.c

@@ -124,7 +124,7 @@ ox_parse(char *xml, ParseCallbacks pcb, char **endp, Options options, Err err) {
     /* initialize parse info */
     helper_stack_init(&pi.helpers);
     // Protect against GC
-    wrap = rb_data_object_wrap(rb_cObject, &pi, mark_pi_cb, NULL);
+    wrap = Data_Wrap_Struct(rb_cObject, mark_pi_cb, NULL, &pi);
 
     err_init(&pi.err);
     pi.str = xml;
@@ -617,6 +617,7 @@ read_element(PInfo pi) {
 			    *start = '\0';
 			    break;
 			case NoSkip:
+			case OffSkip:
 			default:
 			    break;
 			}
@@ -745,6 +746,7 @@ read_text(PInfo pi) {
 			}			
 			break;
 		    case NoSkip:
+		    case OffSkip:
 		    default:
 			*b++ = c;
 			break;

data/ext/ox/sax.c

@@ -175,11 +175,7 @@ sax_drive_init(SaxDrive dr, VALUE handler, VALUE io, SaxOptions options) {
     dr->buf.dr = dr;
     stack_init(&dr->stack);
     dr->handler = handler;
-#if HAS_DATA_OBJECT_WRAP
-    dr->value_obj = rb_data_object_wrap(ox_sax_value_class, dr, 0, 0);
-#else
-    dr->value_obj = rb_data_object_alloc(ox_sax_value_class, dr, 0, 0);
-#endif
+    dr->value_obj = Data_Wrap_Struct(ox_sax_value_class, 0, 0, dr);
     rb_gc_register_address(&dr->value_obj);
     dr->options = *options;
     dr->err = 0;
@@ -1157,7 +1153,9 @@ read_text(SaxDrive dr) {
 	int	isEnd = ('/' == buf_get(&dr->buf));
 
 	buf_backup(&dr->buf);
-	if (NoSkip == dr->options.skip && dr->has.text && !isEnd) {
+	if (dr->has.text &&
+	    ((NoSkip == dr->options.skip && !isEnd) ||
+	     (OffSkip == dr->options.skip))) {
 	    args[0] = rb_str_new2(dr->buf.str);
 #if HAS_ENCODING_SUPPORT
 	    if (0 != dr->encoding) {

data/lib/ox/version.rb

@@ -1,5 +1,5 @@
 
 module Ox
   # Current version of the module. 
-  VERSION = '2.7.0'
+  VERSION = '2.8.0'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ox
 version: !ruby/object:Gem::Version
-  version: 2.7.0
+  version: 2.8.0
 platform: ruby
 authors:
 - Peter Ohler
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2017-09-18 00:00:00.000000000 Z
+date: 2017-09-22 00:00:00.000000000 Z
 dependencies: []
 description: "A fast XML parser and object serializer that uses only standard C lib.\n
   \           \nOptimized XML (Ox), as the name implies was written to provide speed
@@ -35,6 +35,7 @@ files:
 - ext/ox/cache8.c
 - ext/ox/cache8.h
 - ext/ox/dump.c
+- ext/ox/encode.h
 - ext/ox/err.c
 - ext/ox/err.h
 - ext/ox/extconf.rb